7.4 Intracluster Enhanced Location CAC

This topic describes how Enhanced Location CAC works in an intracluster deployment.

Network Model Components

  • Location
    1. Represents a physical location or site
    2. Applied to devices (endpoints, trunks, gateways)
    3. Can be configured with bandwidth limits for calls within the location
  • Link
    1. Connection between two locations
    2. Configured with bandwidth limits and weight
  • Weight
    1. Determines the cost of a link
    2. Used for selection of effective path
  • Effective path
    1. Best path between a pair of locations
    2. Selected by lowest cumulative weight

The intracluster Enhanced Location CAC network model consists of the following components:

  • Location: This component represents a physical location or site. A location is applied to devices such as endpoints, trunks, and media termination points. Depending on the device, the location can be applied directly to the device or via a device pool. With Enhanced Location CAC, a location can be configured with bandwidth limits for calls within a location. This configuration was not possible with Location CAC. The bandwidth limits that were configured with Location CAC applied to all calls coming into and going out of a location.
  • Link: A link connects a pair of locations. Each link is configured with bandwidth limits (one per bandwidth pool: audio, video, and immersive video) and a weight.
  • Weight: The weight determines the cost of a link. Weights are used for selection of the effective path.
  • Effective path: A path is a sequence of links and intermediate locations that connect a pair of locations. The effective path is the “best” path between any given pair of locations. “Best” means that it is the path that has the lowest cumulative weight of all links that comprise the path.

Network Model Example

This section shows an example of an Enhanced Location CAC network model.

  • Network model consists of 9 locations and 12 links.
    1. Each link has a weight assigned.
    2. Each link has bandwidth limits (one per call type) assigned.
  • Call between location 2 and location 9 is placed.
    1. Required bandwidth is determined based on endpoint and region configuration.
    2. Each link along the effective path must have sufficient bandwidth available.
    3. If sufficient bandwidth is available on each link, the call is admitted and available bandwidth of each link is reduced by the bandwidth of the new call.

In the example, the network model consists of 9 locations and 12 links. Each link has an assigned weight. In addition (not shown in the figure), each link would be configured with bandwidth limits (one per type: audio, video, and immersive video).

Cisco Unified Communications Manager calculates the cumulative weight for each possible path between each possible pair of locations. The path with the lowest cumulative weight is the effective path for the given pair of locations. For example, the effective path between location 2 and location 9 is the following: location 2 – location 1 – location 5 – location 6 – location 7 – location 9. The cumulative weight for this path is 100. All other paths between location 2 and location 9 have a higher cumulative weight and are therefore not considered for calls between these two locations.

When a call is placed from location 2 to location 9, the required bandwidth is determined based on endpoint and region configuration codec to be used. Each link along the effective path must have sufficient bandwidth available. If sufficient bandwidth is available, then the call is admitted and the available bandwidth of each link is reduced by the bandwidth that is required by the call.

When designing the network model, it is important that RTP packets are routed according to the chosen effective paths. If they are not routed in this manner, Enhanced Location CAC may admit calls between two locations because, according to the network model, there is enough bandwidth available. But in reality, the IP packets may take a different path, which does not have enough bandwidth provisioned.

The effective path between locations 8 and 5 is via the link that directly interconnects these two locations (the weight is 70). The path via location 6 has a higher weight (80). If the routing protocol metrics prefer the path via location 6 because the two used links have higher bandwidth than the direct link between locations 8 and 5, then all calls that are placed between locations 8 and 5 would deduct bandwidth from the link that interconnects locations 8 and 5. The bandwidth of the links that are actually used for RTP packet transmission would not be reduced by these calls, and the links may become overloaded as a result.

Location Bandwidth Manager

This section describes the role of the LBM service in an Enhanced Location CAC deployment.

  • LBM
    1. Cisco Unified Communications Manager feature service
    2. Responsible for locations, paths, and links assembly
    3. Selects effective paths and performs bandwidth management
    4. Replicates data to all other LBMs within the same cluster
  • LBM Group
    1. Pair of active and standby LBM servers
    2. Assigned to Cisco Unified CM servers (call control servers)
    3. Controls where a Cisco Unified CM sends bandwidth requests:
      1. If LBM group is applied to Cisco Unified CM, active and standby (if configured) LBM servers are used per group definition.
      2. If no group is applied or neither active nor standby LBM server is reachable, local LBM is used (if LBM feature service is coresidentially enabled on Cisco Unified CM server)
      3. If no LBM is reachable, service parameter Call Treatment When No LBM Available is used to admit or deny call.

LBM is a function that is enabled as a Cisco Unified Communications Manager feature service. The service must be activated on at least one server in a cluster. It is not recommended to activate LBM on the publisher. LBM should run on a subscriber that is also used for call processing (that is, the Cisco CallManager service is activated on the server) or on a dedicated server as a standalone service.

For redundancy, it is recommended to have more than one LBM per cluster.

The LBM service is responsible for locations, paths, and links. It determines the effective path for each pair of locations. LBM further performs bandwidth management, which means the LBM service keeps track of maximum and currently available bandwidth per link and replies to CAC queries that are sent from the Cisco CallManager service.

If more than one LBM exists in a cluster, each LBM replicates data to all other LBMs of the same cluster. Replication within a cluster occurs automatically.

A Cisco Unified Communications Manager instance is an entity in the Cisco Unified Communications Manager Administration GUI. This entity represents a Cisco Unified Communications Manager server that has the Cisco CallManager service activated. The Cisco CallManager service performs call-processing functions. Each Cisco Unified Communications Manager instance sends CAC queries to the LBM service. If multiple LBMs exist in a cluster, LBM groups can be configured to control which Cisco Unified Communications Manager instance uses which LBMs within the cluster. Thus, LBM groups provide redundancy (up to two LBMs can be configured per group) and load-sharing (different Cisco Unified Communications Manager instances can refer to different LBMs).

Each LBM group is configured with an active and optionally with a standby LBM server. An LBM server can be a member of multiple groups. Each Cisco Unified Communications Manager instance can have one LBM group applied to it. The same LBM group can be applied to multiple Cisco Unified Communications Manager instances. Using an LBM group configuration is recommended but not mandatory.

Procedure

Procedure

Enhanced Location CAC is enabled by configuring locations with bandwidth limits for intralocation calls or by configuring links and, if applicable, intermediate locations between the originating and terminating locations. Each link is configured with bandwidth limits and a weight. When a Cisco Unified Communications Manager instance wants to set up a call within a location or between two locations that are configured with Enhanced Location CAC, the following steps occur:

  1. If the Cisco Unified Communications Manager instance that wants to route the call has an LBM group that is applied, then the CAC request is sent to the active member of the LBM group. If the active LBM is not reachable, the Cisco Unified Communications Manager instance attempts to query the standby LBM server.

  1. If no LBM group is applied to the Cisco Unified Communications Manager instance or if neither the active nor the standby LBM server responds to the request, then the Cisco Unified Communications Manager instance will fall back to the local LBM service, if configured.

Note

An LBM service is local when it is running on the same server as the Cisco CallManager service.

  1. If no local LBM is present, then the Cisco CallManager clusterwide service parameter Call Treatment, when no LBM is available, determines whether the call should be admitted or denied.

LBM Example

This section provides an example of the interaction between Cisco Unified Communications Manager instances and LBMs.

The figure shows a cluster with four nodes. Nodes 1 and 3 are both running the Cisco CallManager service (that is, it is an operational Cisco Unified Communications Manager instance) and the LBM service. Nodes 2 and 4 both provide call processing functionality only. Only the Cisco CallManager feature service is activated on nodes 2 and 4.

Two LBM groups are configured:

  • LBM-Group_A is configured with LBM_3 as the only group member (active member).
  • LBM-Group_B is configured with LBM_1 as the active member and LBM_2 as the standby member.

LBM-Group_A is applied to the Cisco Unified Communications Manager instance that is running on Node_1, and LBM-Group_B is applied to the Cisco Unified Communications Manager instance that is running on Node_2.

This configuration results in the following priorities (from high to low) for managing Enhanced Location CAC requests:

  • Cisco Unified Communications Manager instance 1: LBM_3 (active and only member of the applied LBM group), LBM_1 (local LBM), and service parameter
  • Cisco Unified Communications Manager instance 2: LBM_1 (active member of the applied LBM group), LBM_3 (standby member of applied LBM group), and service parameter
  • Cisco Unified Communications Manager instance 3: LBM_3 (local LBM) and service parameter
  • Cisco Unified Communications Manager instance 4: Service parameter

Intracluster Enhanced Location CAC: Configuration Procedure

This section describes the steps that are required to configure Enhanced Location CAC.

Enable Enhanced Location CAC:

  1. Activate Cisco LBM feature service.
  2. Configure locations:
    1. Add location.
    2. Configure links to other locations (weight and bandwidth limits).
    3. Optionally configure bandwidth limits within the location.
  3. Configure service parameter.
  4. Apply location to devices.

Control LBM usage:

  1. Configure LBM Group
    1. Add LBM group.
    2. Choose active and standby members.
  2. Apply LBM group to Cisco Unified CM.

The steps can be grouped into two main tasks: enabling Enhanced Location CAC and optionally controlling which Cisco Unified Communications Manager instance should use which LBMs.